Connecting Waterpeople
Premium content

When implementing AI-based technology, the best design makes complex things look simple

Today, Thursday, September 28th, Smart Water Magazine in collaboration with Siemens organized a live webinar titled Harnessing AI for optimization and sustainability in the water and wastewater industry. During the event, Artificial Intelligence (AI) was introduced as a multi-layered solution that holds tremendous potential for water utilities to achieve their sustainability goals

Currently, AI-based technologies are being utilized to optimize diverse processes within water treatment facilities. A prominent pioneer in this domain is Siemens, actively leveraging AI across design, engineering, operations, and maintenance domains. For example, AI plays a critical role in identifying asset wear or malfunctions, like those in pumps and pipelines, and in accurately detecting leaks within water networks. Furthermore, AI-driven predictive modeling aids in accelerating the transition to net zero by enhancing network efficiency and reducing the environmental footprint of water infrastructure.

If you missed the webinar, you may now watch it on-demand.

In this webinar, experts from Siemens explained what Artificial Intelligence (AI) is and how it can benefit the water and wastewater utility and included a live demo of the SIWA LeakPlus solution

The webinar was moderated by Cristina Novo, Technical Editor at Smart Water Magazine, who during her introductory presentation highlighted the endless possibilities Artificial Intelligence provides to water businesses, but reminded the attendees that having a clear strategy for its implementation is key to ensure success.

Cristina Novo, then, introduced Adam Cartwright, Industry Strategy Director – Water Industry, Siemens UK, who began his 30-minute presentation by introducing the concept of Artificial Intelligence (AI), reminding the attendees that it is a wide-ranging branch of computer science that holds incredible potential, but “what you need for a successful implementation is a compelling and effective AI product to actually bring value to your business.”

About Siemens

After briefly introducing AI, Cartwright spoke about his company Siemens, highlighting that “it is a huge, large and diverse engineering service provider.” Although many in the water industry might not know it, Siemens is “one of the largest industrial software companies in the world and has spent over 11 billion dollars over the last ten years acquiring a wide range of different software companies in a wide range of aspects.”

What is Artificial Intelligence?

There’s a large number of academic papers that are taking these approaches and tentatively testing them at various scales, typically pilot scales to see how they work

Next, Adam Cartwright dove deep into the heart of the subject of this webinar: what is AI? He said that nowadays everyone is talking about this interdisciplinary science, whether it is at home with ChatGPT or in any business, or engineers like himself asking how it can be applied to make the most of the solution. However, he warned that AI has “many layers” mimicking how a human might operate and even how it may learn. This aspect of AI is one of the differentiating aspects between this technology and machine learning, for example. He then explained what “deep learning” is, saying it is a subset of machine learning. When speaking of Artificial Intelligence, terms like supervised learning, unsupervised learning or reinforcement learning will generally come up explained Cartwright. “Each of these are different methods that can be applied by AI technologies.” These different methods are used in the water and wastewater sector, for example, to detect leaks or for the optimization of dissolved oxygen in a wastewater treatment process. 

Siemens currently has around 1,400 data scientists who are working in the field of AI across all the different industries. In the water sector, Siemens offers various solutions for example for leak detection and predicting usage which would provide an understanding of the water demand. Cartwright then explained that the collection of data is critically important for AI and machine learning, as it is the input that will then be used to drive the algorithms that drive AI. However, he warns that in the water industry, we have to be really careful and sensitive about data privacy and having really clear demarcations between what can be attributed to an individual and what is purely industrial data.”

In one of our PLCs, for example, in one hour, we can process as much data as is currently existing in all Netflix movies. But the vast majority of the PLC data would give no insight whatsoever. It’d be lacking in features

Risk and failure costs were then discussed by Adam highlighting that usually in a consumer environment, people might do A and B testing; however, this cannot be applied to the water and wastewater sector, as caution is necessary as any actions taken could impact the people and environment you are serving.

How to deploy AI at a large scale?

Afterwards, the speaker highlighted some of the lessons learned by Siemens throughout the years about bringing AI technologies to deployment scale. The first thing companies must do is a pilot project, he said, comically highlighting that “the water industry has more pilots than an airline.” He mentioned that there are several barriers to overcome, such as not having enough data, the culture and how to use these technologies, and how to link the AI application into the wider OT environment. The first lesson he said was not to start off with the technology, but with the value it may bring and making sure you have a clear roadmap from pilot to scale and have an idea of what success would look like at scale. It is important not to “just get excited about the technology.” The second lesson is ensuring the technology is reliable, user-friendly and designed to fit into your broader ecosystem. The third lesson he mentioned is making sure you can interpret the data by using the right tools, otherwise, it can all become a bit overwhelming.

Live demo of SIWA LeakPlus

Following Cartwright’s thorough presentation on AI and how water companies can deploy it at large scale, Holger Hanss, Business Developer of Digital Apps - Water Industry, Siemens AG, introduced Siemens’ smart applications for greater efficiency; SIWA to be precise. A family of applications developed especially for the needs of the water industry. Thanks to SIWA applications, Hanss explained that operators could optimize energy efficiency, avoid water losses, reduce contamination of water bodies and take preventive maintenance measures. He then went on to give a fascinating live demo of SIWA LeakPlus, an AI-based leakage detection for water distribution networks that not only detects leaks and bursts, but also pressure drops, asset failures, water quality problems and sensor failures.

After the presentations, there was a 15-minute Q&A session in which all the attendees were able to share their questions and doubts about Artificial Intelligence (AI) and how plant operators can successfully implement this technology in their operations. Adam Cartwright and Holger Hanss answered questions regarding, for example, the importance of the quality of data compared to the volume of data, the proof of value and trust issues that exist when sharing data in the water industry to solve the major environmental challenges we are facing.